Ethereum and Base have abandoned efforts to align on a single account abstraction standard, leaving two competing proposals, EIP-8130 and EIP-8141, on separate development paths after their joint collaboration broke down last week. Ethlabs researcher Derek Chiang, who tracked the negotiations, said the split became final on Monday, and that wallet developers may soon need to support two different native transaction formats across the two ecosystems. The breakdown is one of the more consequential protocol disputes in the Ethereum ecosystem this year, because it touches the layer every user interacts with: the wallet.
What the two proposals do
Both proposals aim at the same goal: moving account abstraction closer to the protocol level. Today, smart contract wallets on Ethereum are built by wrapping ordinary transactions in complex code, which makes them hard to program and expensive to run. Native support would let wallets offer features like paying gas fees in tokens other than ETH, signing transactions with phone passkeys, batching multiple operations into one transaction and recovering accounts without seed phrases, all without the workarounds that exist today. Those features are widely seen as prerequisites for mainstream adoption, since most people outside the industry will not manage seed phrases no matter how much education the sector produces.EIP-8141, known as Frame Transactions, is the Ethereum L1 track. EIP-8130 is the Base track, backed by the Coinbase-incubated layer 2. The two teams spent months trying to merge the proposals into one standard that both chains would ship. They agreed on the feature list: passkey wallets, flexible authentication, batched transactions, gasless flows. They could not agree on the design underneath those features, which is where standards usually live or die.
“While we identified a number of technical solutions, they all required one side or the other to compromise at least a little bit on their core goals,” Chiang said in a Monday X post.
Why the talks failed
The sticking points map onto the two chains’ different priorities. Ethereum L1 developers optimize for censorship resistance, privacy and security, and they are reluctant to adopt designs that make transaction ordering easier for centralized sequencers to control. Base optimizes for scale, customization and regulatory compliance, and its team sees features like transaction sponsorship and policy controls as table stakes for mainstream users, including the enterprise customers Coinbase courts.Chiang said every technical bridge they built required one side to give up something core. Rather than ship a standard both teams resented, they split. That outcome has precedent in Ethereum’s history, where competing EIPs sometimes coexist until one wins by adoption, but it is rare for two ecosystems this closely linked, with Base settled on the OP stack and anchoring to Ethereum for data availability, to formalize separate wallet standards. The relationship between the two teams remains functional on other fronts, and neither side has framed the split as hostile, but the practical effect is the same regardless of tone.
The fragmentation cost
The practical consequence falls on wallet developers and application teams. If both proposals ship as planned, a wallet that wants to work natively on Ethereum L1 and Base must implement two transaction formats, two signature schemes and two sets of account semantics. That is real engineering cost, and history suggests smaller teams will pick one chain and treat the other as second priority, which deepens the split rather than narrowing it.User experience could suffer in subtle ways. A passkey registered on one chain may not work as expected on the other. Sponsored gas flows may behave differently. Security auditors will need to review two standards instead of one, and tooling like bundlers and paymasters will fragment along the same lines. The account abstraction ecosystem spent three years consolidating around EIP-4337, and this split partially undoes that consolidation at exactly the moment when agent-driven transactions, where software wallets act on behalf of AI systems, are pushing the feature set into production.
What happens next
Neither side has walked back its roadmap. Ethereum researchers are expected to advance EIP-8141 through the protocol upgrade process, where it would compete for inclusion in a future hard fork alongside the EIPs the Ethereum Foundation has already flagged for its next upgrade. Base will likely move faster, since layer 2 networks control their own upgrade schedules and can ship EIP-8130 without ecosystem-wide consensus, and speed is one of the ways L2s compete for developers.The divergence also raises questions for other layer 2s. Arbitrum, Optimism and zkSync chains face the same choice: adopt one of the two standards, support both, or build a third path. If the major L2s fragment, wallet support matrices become the new compatibility headache for the industry, similar to the browser wars of an earlier era. Some developers are already calling for an umbrella abstraction layer that maps between the two formats, though that adds complexity rather than removing it, and abstraction layers have a way of becoming standards of their own.For users, the near-term impact is limited. Existing wallets keep working, and most consumer flows run through custodial or embedded wallets that abstract these details away. The cost lands on the builders, and it lands now, during the design phase, when switching standards is still cheap. By this time next year, the split will either have produced two mature ecosystems or forced a reunion under adoption pressure, and the teams behind both proposals have left the door open to revisiting alignment once real implementations exist and the tradeoffs can be measured rather than argued.